Quantifying heterogeneity in the relationship between R&D intensity and growth at innovative Japanese firms: A quantile regression approach

Abstract

This paper focuses on innovative manufacturing firms in Japan in 2009–2020 and evaluates differences in the relationship between R&D intensity and firm growth. We use a longitudinal version of the conditional quantile regression model to estimate the augmented Gibrat’s law equation for each of four innovative industries: chemicals and allied products; electronic and other electrical equipment; industrial and commercial machinery and computer equipment; and transportation equipment. The analysis reveals statistical differences in estimated coefficients for R&D intensity across low, median and high-growth firms within each industry and across pairs of industries. The results imply the presence of different patterns of R&D effectiveness which are discussed in the light of R&D management drawing on the experience of Sony and other fast-growing Japanese electronics firms. We also discover heterogeneity in the impact on growth of the age and size of firms.